LMSYS Chatbot Arena: Human Preference Prediction 🤖
Using DeBERTa-v3-base, this study predicts human preferences between two chatbot responses based on prompts and responses.
ChingHui
36 views • Dec 1, 2025
About this video
本研究以 Kaggle「LMSYS Chatbot Arena」競賽資料集為基礎，實作一套以 DeBERTa-v3-base 為主體，用於預測人類對兩個大型語言模型回應的偏好。本競賽的主要任務是根據 prompt、response A 與 response B，判斷使用者偏向選擇 A、B 或兩者並列（tie）。
Video Information
Views
36
Likes
1
Duration
14:19
Published
Dec 1, 2025
Related Trending Topics
LIVE TRENDSRelated trending topics. Click any trend to explore more videos.